Processing facilities and other facilities routinely include tanks for storing liquid materials and other materials. For example, storage tanks are routinely used in tank farm facilities and other storage facilities to store oil or other materials. As another example, oil tankers and other transport vessels routinely include numerous tanks storing oil or other materials.
Often times, it is necessary or desirable to measure the amount of material stored in a tank. This may be useful, for example, during loading of material into the tank or unloading of material from the tank. As particular examples, “custody transfers” and “weights and measures of oil” often require highly accurate measurements from level gauging instruments installed on the roof of a tank. In bulk storage tanks, an error of one millimeter in a level reading can correspond to several cubic meters of volumetric error. This can result in losses of thousands of dollars for one or more parties.
One approach to measuring the amount of material in a tank involves the use of radar measurements. In this approach, electromagnetic signals are transmitted towards and reflected off the surface of material in a tank. In radar measurements, one common signal processing task is to detect and identify the reflection of the electromagnetic signals from a desired target, which in this case represents the surface of the material in the tank. In practice, however, the surface of the material is not the only thing that can reflect electromagnetic signals within a tank. Other objects inside a tank (such as the tank's bottom, agitators, heating coils, and ladders) may cause reflections as well. As a result, the task of target detection quickly becomes the task of identifying and tracking the target's reflection among various other reflections and interferences.
If the target's reflection can be separated from other reflections, it is relatively straightforward to detect and track the target. Unfortunately, this often does not occur in real-life situations, such as when the material's surface is close to an antenna, the tank roof, or the tank bottom or when obstructions are present. In these situations, mutual interactions or interferences between reflections can occur. As a consequence, the position and amplitude of the target's refection can be distorted considerably by destructive and constructive interferences.
In conventional systems, a target's refection can often be lost, and wrong level information can be provided to tank operators. This can lead to various problems, such as the overfilling of a tank. In the case of automated tank loading, the likelihood of tank overfilling increases even more. One approach to avoiding these problems is to use fuzzy logic scoring systems to track a limited number of reflection peaks (such as five to ten peaks). Another approach is to use empty tank spectra, or spectra captured when a tank is empty, to map reflections. However, these approaches often require pre-settings in software and historical data saved in memory. These approaches also often increase the complexity of commissioning a gauging system. In addition, empty tank spectra cannot be updated without emptying a tank, which results in lost revenue.